Transcript B - Simon Fraser University

Database Systems II
Index Structures
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
111
Introduction
We have discussed the organization of records
in secondary storage blocks.
Records have an address, either logical or
physical.
But SQL queries reference attribute values, not
record addresses.
SELECT * FROM R WHERE a=10;
How to find the records that have certain
specified attribute values?
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
112
Introduction
value
value
?
index
recordID1
value
recordID2
...
value
blocks
holding
records
matching
records
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
113
Index-Structure Basics
Storage structures consist of files.
Data files store, e.g., the records of a relation.
Search key: one or more attributes for which we
want to be able to search efficiently.
Index file over a data file for some search key
associates search key values with pointers to
(recordID = rid) data file records that have this
value.
Sequential file: records sorted according to their
primary key.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
114
Index-Structure Basics
Sequential File
10
20
30
40
50
60
70
80
90
100
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
115
Index-Structure Basics
Three alternatives for data entries k*:
- record with key value k
- <k, rid of record with search key value k>
- <k, list of rids of records with search key k>
Choice is orthogonal to the indexing technique
used to locate entries k*
Two major indexing techniques:
- tree-structures
- hash tables.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
116
Index-Structure Basics
Dense index: one index entry for every record in
the data file.
Sparse index: index entries only for some of the
record in the data file. Typically, one entry per
block of the data file.
Primary index: determines the location of data file
records, i.e. order of index entries same as order
of data records.
Secondary index does not determine data location.
Can only have one primary index, but multiple
secondary indexes.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
117
Index-Structure Basics
Dense Index
10
20
30
40
50
60
70
80
90
100
110
120
Sequential File
10
20
30
40
50
60
70
80
90
100
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
118
Index-Structure Basics
Sparse Index
10
30
50
70
90
110
130
150
170
190
210
230
Sequential File
10
20
30
40
50
60
70
80
90
100
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
119
Index-Structure Basics
Duplicate key values
– sparse index
– data entry for first new key from block
10
20
30
40
10
10
10
20
20
30
30
30
40
45
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
120
Index-Structure Basics
Sparse index:
- requires less index space per record,
- can keep more of index in memory,
- needed for secondary indexes.
Dense index:
- can tell if any record exists without accessing
data file,
- better for insertions.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
121
Index-Structure Basics
Index file can become very large, e.g. at least
one tenth of data file size for records with ten
attributes of same length.
To speed-up index access, add a second index
level on top of the first index level, a third level
on top of the second one, . . .
First level can be dense, other levels are sparse.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
122
Index-Structure Basics
Sequential File
Sparse 2nd level
10
90
170
250
330
410
490
570
10
30
50
70
90
110
130
150
170
190
210
230
10
20
30
40
50
60
70
80
90
100
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
123
Index-Structure Basics
Index structure needs to support Equality
Queries and Range Queries.
Equality query: one attribute value specified, e.g.
docID = 100, or age = 18.
Range query: attribute range specified, e.g.
30 <= age <= 40.
Index structures must also support DB
modifications, i.e. insertions, deletions and
updates.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
124
ISAM
ISAM = Index Sequential Access Method
Hierarchy of index files (tree structure)
Non-leaf
blocks
Leaf
blocks
Overflow
block
Primary blocks
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
125
ISAM
Leaf blocks contain data entries.
Non-leaf blocks contain pairs (ki,pi), where ki is
a search key value and pi a pointer to the (first
of the) records with that search key value.
P
0
K
1
P
1
K 2
P
2
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
K m
Pm
126
ISAM
File Creation
Leaf (data) blocks allocated sequentially, sorted
by search key.
Then non-leaf blocks allocated, then space for
overflow blocks.
Index entries: <search key value, block id>;
they ‘direct’ search for data entries, which are
in leaf blocks.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
127
ISAM
Example
Root
40
10*
15*
20
33
20*
27*
51
33*
37*
40*
46*
51*
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
63
55*
63*
97*
128
ISAM
Index Operations
Search
Start at root; use key comparisons to go to
leaf.
Insert
Find leaf data entry belongs to, and put it
there.
Delete
Find and remove from leaf; if empty overflow
block, de-allocate.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
129
ISAM
Example
Root
40
Index
blocks
20
33
20*
27*
51
63
Primary
Leaf
10*
15*
33*
37*
40*
46*
48*
41*
51*
55*
63*
97*
blocks
Overflow
23*
blocks
42*
After inserting 23*, 48*, 41*, 42*
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
130
ISAM
Example
Root
40
10*
15*
20
33
20*
27*
23*
51
33*
37*
40*
46*
48*
41*
63
55*
63*
After deleting 42*, 51*, 97*.
51* appears in index level, but not in leaf level.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
131
ISAM
Discussion
Inserts / deletes affect only leaf pages.
 static tree structure
Tree can degenerate into a linear list of
overflow blocks.
In this case, ISAM looses all advantages
compared to a simple, non-hierarchical index
file.
Can we maintain a balanced tree structure
dynamically under insertions / deletions?
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
132
B-Trees
Introduction
Tree node corresponds to block.
B-trees are balanced, i.e. all leaves at same
level. This guarantees efficient access.
B-trees guarantee minimum space utilization.
n (order): maximum number of keys per node,
minimum number of keys is roughly n/2.
Exception: root may have one key only.
m + 1 pointers in node, m actual number of
keys.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
133
B-Trees
Introduction
Index Entries
(inner nodes)
Data Entries
(leaf nodes)
 leaf nodes are linked in sequential order
 this B-tree variant is normally
referred to as B+-tree
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
134
B-Trees
Introduction
Node format: (p1,k1, . . ., pn,kn,pn+1)
pi: pointer, ki: search key
Node with m pointers has m children and
corresponding sub-trees.
n+1-th index entry has only pointer. At leaf
level, this pointer references the next leaf node.
Search key property: i-th subtree contains data
entries with search key k <ki, i+1-th subtree
contains data entries with search key k >= ki.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
135
B-Trees
Example
n=3
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
180
200
150
156
179
120
130
100
101
110
30
35
3
5
11
30
120
150
180
100
Root
136
B-Trees
95
81
Non-leaf
(inner) node
57
Example
to keys
to keys
to keys
to keys
< 57
57 k<81
81k<95
95
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
137
B-Trees
Example
57
81
95
To record
with key 81
To record
with key 85
Leaf node
To record
with key 57
From non-leaf node
to next leaf
in sequence
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
138
B-Trees
Space utilization
120
150
180
30
30
35
Non-leaf
min. node
3
5
11
full node
Leaf
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
counts even if null
n=3
139
B-Trees
Space utilization
Number of pointers/keys for B-tree
Max Max Min
ptrs keys ptrsdata
Non-leaf
(non-root)
Leaf
(non-root)
Root
Min
keys
n+1
n
(n+1)/2 (n+1)/2- 1
n+1
n
(n+1)/2
(n+1)/2
n+1
n
1
1
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
140
B-Trees
Equality Queries
To search for key k, start from root.
At a given node, find “nearest key” ki and
follow left (pi) or right (pi+1) pointer
depending on comparison of k and ki.
Continue, until leaf node reached.
Explores one path from root to leaf node.
Height of B-tree is O(log n / 2 N )
where N: number of records indexed
 runtime complexity O (log N )
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
141
B-Trees
Insertions
Always insert in corresponding leaf.
Tree grows bottom-up.
Four different cases:
space available in leaf,
leaf overflow,
non-leaf overflow,
new root.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
142
B-Trees
Insertions
100
Space available in leaf
Insert key 32
30
31
32
3
5
11
30
n=3
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
143
B-Trees
Insertions
30
n=3
Insert key 7
30
31
3
57
11
3
5
7
split overflowing node into
two of (almost) same size
and copy middle (separating)
key to father node
100
Leaf overflow
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
144
B-Trees
Insertions
split overflowing node
and push middle key
up to father node
180
120
150
180
160
100
Non-leaf overflow
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
180
200
160
179
150
156
179
Insert key 160
145
B-Trees
Insertions
New root
40
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
40
45
20
25
10
12
1
2
3
10
20
30
Insert key 45
30
32
40
30
new root
split can propagate
up to the root and
result in new root
146
B-Trees
Deletions
Locate corresponding leaf node.
Delete specified entry.
Four different cases:
Leaf node has still enough entries,
Coalesce with neighbor (sibling),
Re-distribute keys,
Coalesce or re-distribute at non-leaf.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
147
B-Trees
Deletions
Coalesce with neighbor (sibling)
if node underflows
and sibling has enough
space, coalesce
the two nodes
n=4
40
50
10
20
30
40
10
40
100
Delete key 50
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
148
B-Trees
Deletions
Redistribute keys
if node underflows
and sibling has extra
entry, re-distribute
entries of the
two nodes
n=4
35
40
50
10
20
30
35
10
40 35
100
Delete key 50
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
149
B-Trees
Deletions
Non-leaf coalesce
25
Delete key 37
new root
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
40
45
30
37
30
40
25
26
30
20
22
10
14
1
3
10
20
25
40
n=4
150
B-Trees
B-Trees in Practice
Often, coalescing is not implemented.
It is too hard and typically does not gain a lot
of performance.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
151
B-Trees
B-Trees in Practice
Typical order: 200, typical space utilization:
67%. I.e., average fanout = 133.
Typical capacities:
Height 4: 1334 = 312,900,700 records,
Height 3: 1333 = 2,352,637 records.
Can often hold top levels in buffer pool:
Level 1 =
Level 2 =
1 blocks =
133 blocks =
8 Kbytes,
1 Mbyte,
Level 3 = 17,689 blocks = 133 Mbytes.
 O(1) processing for equality queries
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
152
B-Trees
B-Trees in Practice
Order (n) concept replaced by physical space
criterion in practice (‘at least half-full’).
Inner nodes can typically hold many more
entries than leaf nodes.
Variable sized records and search keys mean
different nodes will contain different numbers
of entries.
Even with fixed length fields, multiple records
with the same search key value (duplicates) can
lead to variable-sized data entries.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
153
Hash Tables
Introduction
Tree-based index structures map search key
values to record addresses via a tree structure.
Hash tables perform the same mapping via a
hash function, which computes the record
address.
Search key K
Hash function h
h( K )  {0,..., B  1}
B: number of buckets.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
154
Hash Tables
Introduction
Good hash function should have the following
property: expected number of keys the same
(similar) for all buckets.
This is difficult to accomplish for search keys
that have a highly skewed distribution, e.g.
names.
Common hash function
K = ‘x1 x2 … xn’ n byte character string
h( K )  ( x1  ...  xn ) MOD B
B often chosen as prime number
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
155
Hash Tables
Secondary-Storage Hash Tables
Bucket: collection of blocks.
Initially, bucket consists of one block.
Records hashed to b are stored in bucket b.
If bucket capacity exceeded, link chain of
overflow buckets.
Assume that address of first block of bucket i
can be computed given i.
E.g., main memory array of pointers to blocks.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
156
Hash Tables
Secondary-Storage Hash Tables
Hash tables can perform their mapping directly
.
or indirectly.
.
.
records
(1) key  h(key)
.
.
.
(2) key  h(key)
key 1
record
Index
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
157
Hash Tables
Insertions
To insert record with search key K.
Compute h(K) = i.
Insert record into first block of bucket i that
has enough space.
If none of the current blocks has space, add a
new block to the overflow chain, and store
new record there.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
158
Hash Tables
Insertions
INSERT:
h(a) = 1
h(b) = 2
h(c) = 1
h(d) = 0
h(e) = 1
0
1
2
d
a
c
b
e
3
bucket capacity:
2 records
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
159
Hash Tables
Deletions
To delete record with search key K.
Compute h(K) = i.
Locate record(s) with search key K in bucket i.
If possible, move up remaining records within
block.
If possible, move remaining records from
overflow chain to the previous block and deallocate block.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
160
Hash Tables
Deletions
Delete:
e
f
0
a
1
c
2
b
c d
e
3
f
g
d
maybe move
“g” up
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
161
Hash Tables
Queries
To find record(s) with search key K.
Compute h(K) = i.
Locate record(s) with search key K in bucket i,
following the overflow chain.
In the absence of overflow blocks, only one
block I/O necessary, i.e. O(1) runtime.
This is (much) better than B-trees.
But hash tables do not support range queries!
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
162
Hash Tables
Queries
In order to keep overflow chains short, keep
space utilization between 50% and 80%.
# keys in bucket b
space utilizatio n (b) 
# keys that fit in b
If space utilization < 50%: waste space.
If space utilization > 80%: overflow chains
become significant.
Depends on hash function and on bucket
capacity.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
163
Hash Tables
So far, only static hash tables, i.e. the number B of
buckets never changes.
With growing number of records, space
utilization cannot be kept in the desired range.
Dynamic hash tables adapt B to the actual number
of records stored.
Goal: approximately one block per bucket.
Two dynamic methods:
Extensible Hashing, and
Linear Hashing.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
164
Extensible Hash Tables
Introduction
Add a level of indirection for the buckets, a
directory containing pointers to blocks, one for
each value of the hash function.
h(K)
.
.
.
to bucket
.
.
.
Size of directory doubles in each growth step.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
165
Extensible Hash Tables
Introduction
Several buckets can share a data block, if they
do not contain too many records.
Hash function computes sequences of k bits,
but bucket numbers use only the i first of
these bits. i is the level of the hash table.
k
00110101
size of directory  2 ,
initially i  0
i
i, grows over time
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
166
Extensible Hash Tables
Insertions
To insert record with search key K.
Compute h(K) and take its first i bits. Global
level i is part of the data structure.
Retrieve the corresponding directory entry.
Follow that pointer leading to block b. b has a
local level j <= i.
If b has enough space, insert record there.
Otherwise, split b into two blocks.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
167
Extensible Hash Tables
Insertions
If j < i, distribute records in b based on (j+1)st
bit of h(K): if 0, old block b, if 1 new block b’.
Increment the local level of b and b’ by one.
Adjust the pointer in the directory that
pointed to b but must now point to b’.
If j = i, first increment i by one. Double the
directory size and duplicate all entries.
Proceed as in case j < i.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
168
Extensible Hash Tables
Example
i=
1
1
0001
i=2
00
01
10
1 2
1001
1010 1100
Insert 1010
1 2
1100
11
New directory
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
169
Extensible Hash Tables
i= 2
00
01
10
11
Insert:
0111
0000
2
0000
0001
1 2
0001 0111
0111
2
1001
1010
2
1100
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
170
Extensible Hash Tables
i= 2
00
0000 2
0001
i=3
0111 2
01
10
11
Insert:
1001
1001 3
1001
1010 1001 2 3
1010
1100 2
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
000
001
010
011
100
101
110
111
171
Extensible Hash Tables
Overflow Chains
May still need
overflow chains
in the presence
of too many
duplicates of
hash values.
insert 1100
1
1101
1100
Split does not help if all
entries belong to same
of two resulting blocks!
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
if we split:
2
2
1100
1100
172
Extensible Hash Tables
Overflow Chains
insert 1100
1
1101
1100
add overflow block:
1
1101
1101
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
1100
173
Extensible Hash Tables
Deletions
To delete record with search key K.
Using the directory, locate corresponding
block b and delete record from there.
If possible, merge block b with “buddy” block
b’ and adjust the directory pointers to b and b’.
If possible, halve the directory.
 reverse insertion procedure
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
174
Extensible Hash Tables
Discussion
Can manage growing number of buckets
without wasting too much space.
Assume that directory fits into main memory.
Never need to access more than one data
block (as long as there are no overflow chains)
for a query.
Doubling the directory is a very expensive
operation. Interrupts other operations and
may require secondary storage.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
175
Linear Hash Tables
Introduction
No directory.
Hash function computes sequences of k bits.
Take only the i last of these bits and interpret
them as bucket number m.
k
00110101
i, grows over time
n: number of last bucket, first number is 0.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
176
Linear Hash Tables
Insertions
If m <= n, store record in bucket m. Otherwise,
store it in bucket number m  2 i 1
If bucket overflows, add overflow block.
If space utilization becomes too high, add one
bucket at the end and increment n by 1.
space utilizatio n 
r
, where r  total number of records
(n  1)  c
and c  bucket capacity (number records)
 file grows linearly
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
177
Linear Hash Tables
Insertions
Bucket we add is usually not in the range of
hash keys where an overflow occurred.
When n > 2i, increment i by 1.
i is the number of rounds of doubling the size
of the Linear Hash table.
No need to move entries.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
178
Linear Hash Tables
Example
• insert 0101
0101
1010
0101
1111
00
01
n = 01
k = 4, i = 2
• can have overflow chains!
Future
growth
buckets
10
11
If h(k)[i ]  n, then
look at bucket h(k)[i ]
else, look at bucket h(k)[i ] – 2i -1
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
179
Linear Hash Tables
Example
k = 4, i = 2
0101
1010
00
0101
0101
1111
01
• insert 0101
1010
10
1111
Future
growth
buckets
11
n = 01
10
11
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
180
Linear Hash Tables
Example
k=4
i=2 3
0101
0101
0 00
100
0 01
101
1010
0 10
110
0101
0101
1111
0 11
111
100
101
...
n = 11
100
101
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
181
Linear Hash Tables
Discussion
Can manage growing number of buckets
without wasting too much space.
No directory, i.e. no indirection in access and
no expensive doubling operation.
Significant need for overflow chains, even if
no duplicates among last i bits of hash values.
CMPT 454, Simon Fraser University, Fall 2009, Martin Ester
182